Seveso in the Netherlands: What It Means for Your Site

What Seveso actually is

Seveso is European legislation dealing with the control of major accident hazards involving dangerous substances. It arose from a real accident in an Italian town in 1976, and the directive still bears that town's name. If you want the background, we wrote it up separately: the Seveso disaster and what came out of it.

First, understand this: Seveso is not a certificate. No one gives you a Seveso badge. It is a duty to prevent major accidents and to be able to demonstrate that you are preventing them. These are two different jobs, and the second one causes most of the trouble.

In the Netherlands, these duties reach you through national implementation, now part of the Omgevingswet framework. Many people at plant level still say Brzo out of habit. The vocabulary has shifted, but the substance has not: you are a Seveso-inrichting, you are in a tier, and you owe a written system that matches what actually happens on your site.

Which tier you are in changes what you owe

Your tier is determined by the dangerous substances present against the threshold quantities. There's a lower threshold and an upper threshold. Getting that classification right is the first step, because everything that follows depends on it.

Lower tier

You need a preventiebeleid zware ongevallen, the PBZO-document, and a veiligheidsbeheerssysteem to support it. The VBS is not optional at this tier. The paperwork is less extensive, but the principle is the same.

Upper tier

You need a veiligheidsrapport. The VR covers the installations, the scenarios, and the measures, and it incorporates the PBZO and the VBS. It goes to the bevoegd gezag and must be kept current. There is a periodic review cycle, and there's a separate duty that catches more people out: when something changes on the installation, the VR must reflect that change.

This second duty is where sites quietly drift. A modification happens. It's engineered properly, executed properly, and the veiligheidsrapport is not updated. Six months later, the written system and the physical plant no longer match. Nothing is unsafe. Everything is undemonstrable.

The seven elements are where inspections land

The VBS has seven elements. These form the actual audit surface:

1. Organisation and personnel: who is responsible, who is trained, and can you show it.
2. Identification and evaluation of major hazards: scenarios, bowties, and whether they still align with the installation.
3. Operational control: procedures, work permits, and the checks that are supposed to happen.
4. Management of change: the MOC loop, from proposal through to the documents that must be updated.
5. Planning for emergencies: bedrijfsnoodplan, drills, and evidence they were conducted.
6. Monitoring performance: how you identify your own gaps before someone else does.
7. Audit and review: the internal loop, and what you did with its findings.

Element 4 is the one that generates most of the follow-on work, because management of change is the hinge between the plant and the paperwork. If MOC is weak, elements 2 and 3 become stale on their own.

What inspection actually feels like

Seveso inspections in the Netherlands are conducted jointly by the relevant authorities. Depending on your location and your topic, this includes the bevoegd gezag for environmental matters, such as DCMR in the Rijnmond region, and the Nederlandse Arbeidsinspectie for worker safety.

Two things are worth stating plainly.

First, findings are often small and physical. An eyewash station a few weeks past its expiry date. A checklist that was signed but demonstrably could not have been completed in the time available. These are not deep systemic failures. They are recorded as overtredingen nonetheless, because the standard is demonstrability.

Second, not every written finding is beyond discussion. Inspectors are people working from interpretation, and interpretation can be contested. The useful stance is neither deference nor conflict. It is having the facts assembled well enough that the conversation is about facts rather than about who remembers what. If the engineer who explained a topic three years ago has since left, and nothing was written down, you will lose that discussion on paper even when you were correct on substance.

What to do in the next quarter

Concrete actions, in rough order of payoff:

• Test one document against your own reality. Take a single section of the VR or the PBZO-document and walk the installation it describes. Note every place the text and the plant disagree.
• Check the MOC backlog for document debt. Not the changes themselves. The changes that closed without their VR consequence being processed.
• Find your knowledge single points of failure. Which topics reside in one person's head, and which of those people are nearing retirement.
• Look at your checklists honestly. If a round is four pages and gets two hours, the round is not being done. It is being signed.
• Stop over-referencing in the VR. Every procedure you pull in by reference becomes something you owe the bevoegd gezag. Reference what genuinely belongs there, not everything you can think of.
• Write down the informal answers. The verbal explanations that satisfied an inspector once are worth nothing the second time unless they exist as text.

Where software helps, and where it does not

Software does not make you compliant. It cannot. Seveso duties rest with the operator, and the veiligheidsbeheerssysteem is yours, not your vendor's. Nothing you buy removes the decisions or the risk from your desk.

What tooling can genuinely do is narrow its scope: make the round on the werkvloer take the time it actually needs, make findings turn into tracked corrective actions instead of into a photograph on someone's phone, keep changes connected to the documents they affect, and surface the gap between what you wrote and what you do. That is a smaller claim than most vendors make. It is also the claim that survives an audit.

Generic safety management platforms tend to fall short here for a specific structural reason, which we set out in why generic safety management software falls short for Seveso III companies.

A short, honest next step

Pick one document. One section of one veiligheidsrapport or one PBZO-document. Score it against what is physically true on site today. You will get a number you do not love, and that number is the only honest starting point there is.

If you want a second pair of eyes on that exercise, we do it as a gap analyse against moederdocumenten, and the more of your own documentation we get, the more useful the output is. No obligation to continue after it. Say what you want to look at and we will tell you whether we can help with it.